Google Puts Its Virtual Brain Technology to Work - Technology Review

This summer Google set a new landmark in the field of artificial intelligence with software that learned how to recognize cats, people, and other things simply by watching YouTube videos (see “Self-Taught Software”). That technology, modeled on how brain cells operate, is now being put to work making Google’s products smarter, with speech recognition being the first service to benefit.

Google’s learning software is based on simulating groups of connected brain cells that communicate and influence one another. When such a neural network, as it’s called, is exposed to data, the relationships between different neurons can change. That causes the network to develop the ability to react in certain ways to incoming data of a particular kind—and the network is said to have learned something.

Neural networks have been used for decades in areas where machine learning is applied, such as chess-playing software or face detection. Google’s engineers have found ways to put more computing power behind the approach than was previously possible, creating neural networks that can learn without human assistance and are robust enough to be used commercially, not just as research demonstrations.

I.B.M. Announces Brainy Computer Chip - NYTimes.com
Dharmendra Modha, an I.B.M. researcher, is the leader of the project to create cognitive computer chips.
Since the early days in the 1940s, computers have routinely been described as “brains” — giant brains or mathematical brains or electronic brains. Scientists and engineers often cringed at the distorting simplification, but the popular label stuck.
 
Wait long enough, it seems, and science catches up with the metaphor. The field of “cognitive computing” is making enough progress that the brain analogy is becoming more apt. I.B.M. researchers are announcing on Thursday two working prototype cognitive computer chips.
The chip designs are the result of a three-year project involving I.B.M. and university researchers, supported by the Defense Advanced Research Projects Agency. The academic collaborators are at Columbia University, Cornell University, the University of California, Merced and the University of Wisconsin.
The results to date have been sufficiently encouraging that Darpa is announcing on Thursday that it will commit an additional $21 million to the project, the third round of government funding, which brings the total to $41 million.
The cognitive chips are massively parallel microprocessors that consume very little power. But they also have a fundamentally different design. The two prototype semiconductor cores each has 256 neuronlike nodes. One core is linked to 262,144 synapselike memory modules, while the other is linked to 65,536 such memory synapses.

I.B.M. Announces Brainy Computer Chip - NYTimes.com

Dharmendra Modha, an I.B.M. researcher, is the leader of the project to create cognitive computer chips.

Since the early days in the 1940s, computers have routinely been described as “brains” — giant brains or mathematical brains or electronic brains. Scientists and engineers often cringed at the distorting simplification, but the popular label stuck.

Wait long enough, it seems, and science catches up with the metaphor. The field of “cognitive computing” is making enough progress that the brain analogy is becoming more apt. I.B.M. researchers are announcing on Thursday two working prototype cognitive computer chips.

The chip designs are the result of a three-year project involving I.B.M. and university researchers, supported by the Defense Advanced Research Projects Agency. The academic collaborators are at Columbia University, Cornell University, the University of California, Merced and the University of Wisconsin.

The results to date have been sufficiently encouraging that Darpa is announcing on Thursday that it will commit an additional $21 million to the project, the third round of government funding, which brings the total to $41 million.

The cognitive chips are massively parallel microprocessors that consume very little power. But they also have a fundamentally different design. The two prototype semiconductor cores each has 256 neuronlike nodes. One core is linked to 262,144 synapselike memory modules, while the other is linked to 65,536 such memory synapses.